Raft of optimiser changes
Sergei Trofimovich
slyich at gmail.com
Sat Aug 30 10:38:32 UTC 2014
On Thu, 28 Aug 2014 11:16:03 +0000
Simon Peyton Jones <simonpj at microsoft.com> wrote:
> I've just pushed a bunch of Core-to-Core optimisation changes that have been sitting in my tree for ages. The aggregate effect on nofib is very modest, but they are mostly aimed at corner cases, and consolidation.
>
> Program Size Allocs Runtime Elapsed TotalMem
>
> Min -7.2% -3.1% -7.8% -7.8% -14.8%
>
> Max +5.6% +1.3% +20.0% +19.7% +50.0%
>
> Geometric Mean -0.3% -0.1% +1.7% +1.7% +0.2%
> The runtime increases are spurious - I checked.
> A couple of perf/compiler tests (i.e. GHC's own performance) improve significantly, which is a good sign.
> I have a few more to come but wanted to get this lot out of my hair.
Hello Simon!
The compiler improvements look great!
Although running 'fulltest' one test caught core lint error:
> typecheck/should_compile T7891 [exit code non-0] (hpc,optasm,profasm,optllvm)
It can be reran as:
$ make fulltest THREADS=12 TEST=T7891
The result of optasm run:
=====> T7891(optasm) 3365 of 4096 [0, 0, 0]
cd ./typecheck/should_compile && '/home/slyfox/dev/git/ghc-validate/inplace/bin/ghc-stage2' -fforce-recomp -dcore-lint -dcmm-lint -dno-debug-output -no-user-package-db -rtsopts -fno-ghci-history -c T7891.hs -O -fasm -fno-warn-incomplete-patterns >T7891.comp.stderr 2>&1
Compile failed (status 256) errors were:
*** Core Lint errors : in result of Simplifier ***
<no location info>: Warning:
In the type ‘a_12 -> t_aiE -> t_aiE’
@ a_12 is out of scope
(attached it's complete output)
Thank you!
--
Sergei
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: T7891-optasm-failure.txt
URL: <http://www.haskell.org/pipermail/ghc-devs/attachments/20140830/dfb3046e/attachment-0001.txt>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 181 bytes
Desc: not available
URL: <http://www.haskell.org/pipermail/ghc-devs/attachments/20140830/dfb3046e/attachment-0001.sig>
More information about the ghc-devs
mailing list